Applying data and measurement is an essential part of any marketing strategy. Attention is highly fragmented and consumers are likely exposed to a brand’s campaign in different ways throughout their path to purchase. An attribution plan can serve as a foundation to help advertisers understand the true value of each marketing channel. A long-time partner with Yahoo, Caroline Ballard (Digital Media Director, Starcom USA) shared her perspective on multi-touch attribution.
A few months ago at MediaPost’s Search Insider Summit (#MPSIS), you discussed an important topic for marketers today: measuring the impact of campaigns that reach consumers across different kinds of online experiences and devices, or to use an industry buzzword “multi-touch attribution.” Are there categories that should be more focused on attribution than others?
While I feel strongly that everybody should be thinking about attribution, there’s a big opportunity for categories like CPG. Often, they are measuring growth in awareness; however, as companies are focusing on ROI and bottom-line revenue, it’s important to understand how hard all of their channels are working.
Our partnership with Yahoo on a campaign for one of our CPG clients, Kraft-Heinz, helped us understand how a search campaign with more of a branding focus can still impact sales. This can help us understand the full funnel and ensure we are investing appropriately in the right channels.
What’s the first step toward creating a solution?
If you haven’t already, set up a tagging system and start collecting data. Then, it’s important to work with your agency or internal teams to create an infrastructure to house your data. Files can be significantly large and complex.
Once a marketer is set up to collect this data, what should they do with it?
It’s easy to get lost before you even start looking at attribution insights, especially when you’re ingesting online and offline data. Before you begin analyzing anything, write down your goals. For example, is it most important to link search to offline sales? Are you measuring online revenue but know that 90% of total sales happen in store and aren’t being attributed properly? Identify one or two ideas that are most compelling to your organization and seek answers to those first.
How do you recommend creating an effective attribution model?
One important step is to identify key stakeholders at your company, especially those outside of your core team. This could include IT, revenue management, data scientists, and finance. Think about all the aspects of an attribution model and make sure those teams are on board. If you have access to a data scientist, this can be valuable and neutral party who can analyze data without bias towards one channel or another.
Any final tips for marketers that are just getting started?
Don’t take the first round of results as the only source of truth, but also don’t discount them if they demonstrate that what you’ve been doing isn’t optimal. Use it as a foundation of discussion and be open to interpretation. Like most data insights, attribution is a mix of art and science. You may see the results show the exact opposite of what you thought, and that’s okay, because it means there’s an opportunity to learn. Attribution is a living model that must be tested regularly to find the optimal mix for your brand.